Targeted Upregulation of Pyruvate Dehydrogenase
Kinase (PDK)-4 in Slow-Twitch Skeletal Muscle
Underlies the Stable Modification of the Regulatory
Characteristics of PDK Induced by High-Fat
Feeding
Mark J. Holness, Alexandra Kraus, Robert A. Harris, and Mary C. Sugden
In using Western blot analysis with antibodies raised
against recombinant pyruvate dehydrogenase kinase
(PDK) isoforms PDK2 and PDK4, this study demonstrates selective PDK isoform switching in specific
skeletal muscle types in response to high-fat feeding
that is associated with altered regulation of PDK activity by pyruvate. The administration of a diet high in saturated fats led to stable (~2-fold) increases in PDK
activities in both a typical slow-twitch (soleus [SOL])
muscle and a typical fast-twitch (anterior tibialis [AT])
muscle. Western blot analysis revealed that high-fat
feeding significantly increased (~2-fold; P < 0.001)
PDK4 protein expression in SOL, with a modest
(1.3-fold) increase in PDK2 protein expression. The
relative increase in PDK4 protein expression in SOL was
associated with a 7.6-fold increase in the pyruvate concentration that was required to elicit a 50% active
pyruvate dehydrogenase complex, which indicates a
marked decrease in the sensitivity of PDK to inhibition by pyruvate. In AT muscle, high-fat feeding elicited
comparable (1.5- to 1.7-fold) increases (P < 0.05) in
PDK4 and PDK2 protein expression. Loss of sensitivity
of PDK to inhibition by pyruvate was less marked. The
data suggest that a positive correlation exists between
increases in PDK4 expression and the propensity with
which muscles use lipid-derived fuels as respiratory
substrates rather than with the degree of insulin resistance induced in skeletal muscles by high-fat feeding. In
conclusion, high-fat feeding leads to selective upregulation of PDK4 expression in slow-twitch muscle in
response to high-fat feeding in vivo, which is associated
with a pronounced loss of sensitivity of PDK activity to
acute inhibition by pyruvate. Thus, increased PDK4
From the Department of Diabetes and Metabolic Medicine (M.J.H., A.K.,
M.C.S.), Division of General and Developmental Medicine, St. Bartholomew’s
and the Royal London School of Medicine and Dentistry, Queen Mary and
Westfield College, University of London, London, U.K.; and the Department
of Biochemistry and Molecular Biology (R.A.H.), Indiana University School
of Medicine, Indianapolis, Indiana.
Address correspondence and reprint requests to M.C. Sugden, MA,
DPhil, DSc, Department of Diabetes and Metabolic Medicine, Medical Sciences Bldg., Queen Mary and Westfield College, Mile End Rd., London E1
4NS, U.K. E-mail: [email protected].
Received for publication 11 May 1999 and accepted in revised form
21 January 2000.
[3H]2DG, 2-deoxy-D-[1-3H]glucose; AT, anterior tibialis; ECL, enhanced
chemiluminescence; FA, fatty acid; GUI, glucose utilization index; PDC,
pyruvate dehydrogenase complex; PDHa, active pyruvate dehydrogenase
complex; PDK, pyruvate dehydrogenase kinase; PPAR-, peroxisome proliferator-activated receptor-; Rd, glucose disappearance rate; SOL, soleus;
TBS, Tris-buffered saline; TBST, Tris-buffered saline with Tween.
DIABETES, VOL. 49, MAY 2000
expression may underlie the stable modification of the
regulatory characteristics of PDK observed in slowtwitch muscle in response to high-fat feeding. Diabetes
49:775–781, 2000
R
egulation of the activity of the pyruvate dehydrogenase complex (PDC) is an important component of the regulation of glucose homeostasis.
Activation of PDC promotes glucose disposal,
whereas suppression of PDC activity is crucial to glucose
conservation in starvation when 3 carbon compounds
(including pyruvate) are required for gluconeogenesis to
maintain glycemia. PDC is rendered inactive by phosphorylation of the -subunit of its pyruvate dehydrogenase component by pyruvate dehydrogenase kinase (PDK) (1,2). PDK
activity is itself regulated. Short-term mechanisms of PDK
regulation include suppression of activity by pyruvate generated by glycolysis or from circulating lactate and activation
resulting from high mitochondrial acetyl-CoA/CoA and
NADH/NAD+ concentration ratios that are generated by
increased rates of fatty acid (FA) -oxidation. In addition,
PDK activity can be enhanced by a longer-term mechanism
that is independent of the acute effects of small molecular
weight effectors (3). A total of 4 PDK isoforms have been
identified in mammalian tissues (4,5). Studies with the individual recombinant PDK isoforms expressed in Escherichia
coli have demonstrated important differences in their specific
activities and in their regulatory characteristics (4). In studies in the rat in vivo, stable effects of starvation to enhance
PDK activity have been observed in association with the
upregulation of the protein expression of at least 2 of these
4 PDK isoforms. PDK4 is upregulated in response to starvation in the heart (6). PDK4 is a high–specific activity PDK isoform whose activity is relatively insensitive to suppression
by dichloroacetate (4), which is a highly specific synthetic
inhibitor of PDK that is believed to mimic the effect of the
physiological inhibitor pyruvate. This response to starvation appears to be specific for PDK4 with no change in protein expression of PDK2 (a lower–specific activity pyruvatesensitive isoform) (6). By contrast, PDK2 protein expression is upregulated by starvation both in the liver (7) and in
the kidneys (8). This suggests that the responses of individual PDK isoforms to insulin deficiency are dictated by tissue
type and presumably function.
775
PDK ISOFORM EXPRESSION
The skeletal muscle mass is quantitatively of major importance for glucose disposal (9). In nondiabetic Pima Indians (a
population with a high prevalence of type 2 diabetes associated with obesity), the expression of PDK2 and PDK4 mRNA
in muscle biopsies correlates positively with fasting plasma
insulin levels (an indicator of insulin resistance) and negatively with insulin-mediated glucose uptake (10). This study
suggested that insufficient downregulation of PDK by insulin
in the skeletal muscles of insulin-resistant individuals may be
a cause of increased PDK expression, thus leading to
impaired glucose oxidation followed by increased FA oxidation (10). Recent findings have also implicated a direct role
for FA acting via the peroxisome proliferator-activated receptor- (PPAR-) in signaling increased PDK4 expression in
skeletal muscle (11). However, the skeletal muscle mass displays considerable heterogeneity regarding glucose utilization
(both in the basal [postabsorptive] state and during insulin
stimulation) and the capacity for FA utilization. Under resting conditions, the slow-twitch muscles (e.g., soleus [SOL]
muscle) display higher rates of glucose utilization (12,13)
and greater insulin sensitivity (13) than the fast-twitch muscles (e.g., anterior tibialis [AT] muscle). Furthermore, slowtwitch muscles (which contain higher proportions of oxidative fibers [14]) avidly oxidize FA, and glucose utilization is
markedly reduced in these metabolically active muscles to a
greater degree than the reduction in whole-body glucose disposal after high-fat feeding (15). In contrast, fast-twitch muscles containing mainly glycolytic fibers are little affected by
high-fat feeding (15). PDK2 and PDK4 mRNA expression has
been demonstrated in rat gastrocnemius muscle (4); in the rat,
this muscle displays characteristics of fast-twitch muscle
regarding glucose utilization in the fed and fasted states (12)
and during insulin stimulation (13,16).
Researchers have demonstrated previously that skeletal
muscle PDK activity increases in response to the administration of a diet high in saturated fat and low in carbohydrates
(17). In the present study, we examined to what extent PDK2
and/or PDK4 protein expression in skeletal muscle is altered
and whether individual muscle types respond differently
regarding changes in expression of different PDK isoforms in
response to high–saturated fat feeding, which elicits insulin
resistance in both slow- and fast-twitch muscles. We evaluated
whether changes in the expression of either of these PDK isoforms in slow- and fast-twitch skeletal muscle in vivo correlate positively with their capacity for FA utilization or negatively with changes in skeletal muscle insulin sensitivity
assessed directly in vivo with the 2-deoxy[1-3H]glucose technique in combination with the euglycemic-hyperinsulinemic
clamp. Finally, to assess whether PDK isoform switching, by
virtue of the distinct regulatory properties identified in vitro
with recombinant proteins (4), is of potential physiological
significance for the control of glucose oxidation in vivo, we
investigated whether changes in the PDK isoform profile in
individual skeletal muscles are reflected in altered sensitivity of PDK activity to suppression by pyruvate.
RESEARCH DESIGN AND METHODS
Materials. Human Actrapid insulin was purchased from Novo Nordisk
(Bagsvaerd, Denmark). Kits for determination of plasma insulin concentrations
were purchased from Phadeseph Pharmacia (Uppsala, Sweden). Radiochemicals and enhanced chemiluminescence (ECL) reagents were purchased
from Amersham International (Buckinghamshire, U.K.). Arylamine acetyltransferase was purified from pigeon liver acetone powder (Europa Bio776
products, Ely, Cambridgeshire, U.K.). Other chemicals and biochemicals were
purchased from Bio-Rad (Hemel Hempstead, Hertfordshire, U.K.), Boehringer
Mannheim (Lewes, East Sussex, U.K.), or Sigma (Poole, Dorset, U.K.). Female
Wistar rats were purchased from Charles River (Margate, Kent, U.K.). Diets
and individual components of diets were purchased from Special Diet Services
(Witham, Essex, U.K.).
Rats and diets. Adult female Wistar rats (initial weights 230–250 g) with free
access to food and water were housed in individual cages in a temperature(21 ± 2°C) and light-controlled room (a 12-h light–dark cycle). Age-matched
rats were randomly allocated to 2 groups. The first group was maintained on
standard high-carbohydrate/low-fat rodent laboratory diet (72% digestible
carbohydrate, 20% protein, 8% lipid by energy), whereas the second group was
maintained on a semisynthetic low-carbohydrate high–saturated fat diet (33%
carbohydrate, 20% protein, and 47% lipid by energy) (17) for 28 days. In all of
the experimental protocols, the 2 individual muscles were sampled from the
same animal and were analyzed in parallel.
Euglycemic-hyperinsulinemic clamps. For the euglycemic-hyperinsulinemic
clamp studies, each rat was fitted with 2 chronic indwelling cannulas. One cannula was placed in the right jugular vein, and the other cannula was placed in
the left jugular vein (for infusion and sampling, respectively) under Hypnorm
(Janssen Pharmaceuticals, Oxford, U.K.) (fentanyl citrate [0.315 mg/ml]/
fluanisone [10 mg/ml], 1 ml/kg body wt via intraperitoneal injection) and
Diazepam (Phoenix Pharmaceuticals, Gloucester, U.K.) (5 mg/ml, 1 ml/kg body
wt via intraperitoneal injection) anesthesia at 5–7 days before study (18). On
the day of the experiment, food was withdrawn at the end of the dark (feeding) phase, and the rats were studied in the postabsorptive state at 1400 (i.e.,
6 h after food withdrawal). Euglycemic-hyperinsulinemic clamps were performed in conscious unstressed freely moving rats (18). In brief, a primed continuous intravenous infusion of insulin was administered at a fixed rate
(4.2 mU · kg body wt–1 · min–1) for 2.5 h. This insulin dose was selected on the
basis of previous studies demonstrating that it produces plasma insulin concentrations comparable to those observed after ingesting a carbohydrate-rich
meal (16). A variable rate of glucose infusion was initiated at 1 min after the
start of insulin infusion. Blood was sampled from the right jugular vein at 5to 10-min intervals. A steady state was reached after 60–90 min. Coefficients
of variation of blood glucose concentrations during the hyperinsulinemic
clamp were <12% in all studies. Whole-body glucose kinetics were estimated
in awake unstressed freely moving rats in the basal (postabsorptive) state
and during euglycemia-hyperinsulinemia with a primed (0.5 µCi) continuous
(0.2 µCi · min–1 · rat–1) intravenous infusion of [3-3H]-labeled glucose as previously described (18,19). Whole-body glucose disappearance rates (Rd) were calculated as previously described (18).
In vivo glucose utilization in individual muscles. Estimations of glucose
utilization by individual skeletal muscles in vivo were obtained by measuring
the accumulation of 2-deoxy-D-[1-3H] glucose-6-phosphate in the tissue after the
bolus intravenous injection of tracer amounts (30 µCi) of 2-deoxy-D-[1-3H]glucose ([3H]2DG) in the basal state or during hyperinsulinemia (at 90 min after
the start of the clamp) (12). Blood samples (100 µl) for determination of blood
glucose concentrations and plasma tracer concentrations were obtained at 1,
3, 5, 10, 20, 40, and 60 min after [3H]2DG bolus administration. Throughout the
study, the rats were awake and moving freely with the connecting tubing suspended overhead. At the end of the 60-min study, a final blood sample (500 µl)
was added to a heparinized tube and was immediately centrifuged at 4°C, and
plasma was frozen at –20°C for subsequent insulin determinations. Rats were
killed by the intravenous injection of pentobarbitone (60 mg/kg body wt) via
the right jugular cannula. Individual skeletal muscles (SOL and AT) were
freeze-clamped when locomotor activity had ceased (within 5 s). The fiber profiles (fast oxidative glycolytic:fast glycolytic:slow oxidative) of SOL and AT in
the rat are 0:0:100 and 66:32:2, respectively (14). The freeze-clamped muscles
were stored in liquid nitrogen until analysis as previously described (12,20). No
correction was made for possible discrimination against 2-deoxyglucose versus glucose regarding glucose transport and phosphorylation, and hence rates
of tissue accumulation of 2-deoxy-D-[3H]glucose-6-phosphate are referred to as
glucose utilization index (GUI) values.
Enzyme assays. Active PDC (PDHa) activity was assayed spectrophotometrically by coupling the generation of acetyl-CoA to the acetylation of p-(paminophenylazo)benzene sulfonic acid by arylamine acetyltransferase (21).
PDHa activities are expressed relative to citrate synthase to correct for possible differences in mitochondrial extraction (21). Total PDC activity was
assayed as active PDC after complete activation through the action of endogenous PDC phosphate phosphatase in mitochondria incubated for 15 min in the
absence of respiratory substrate and in the presence of the uncoupler carbonyl
cyanide m-chlorophenylhydrazone (22). PDK activities were determined at
30°C in mitochondrial extracts at a pH of 7.0 by the rate of ATP-dependent inactivation of PDHa (17,23). PDK activities are expressed as first-order rate conDIABETES, VOL. 49, MAY 2000
M.J. HOLNESS AND ASSOCIATES
stants for ATP-dependent PDHa inactivation. To test the effects of pyruvate,
mitochondria were incubated at 30°C in KCl medium (100 mmol/l KCl,
20 mmol/l Tris, 5 mmol/l KH2PO4, 2 mmol/l EGTA, pH 7.4) in the presence of
respiratory substrate (5 mmol/l 2-oxoglutarate/0.5 mmol/l L-malate) together
with the concentrations of pyruvate indicated. Incubations were terminated
by centrifugation after 5 min, and mitochondrial extracts were assayed for
PDHa activity (24).
Western blotting analysis. Mitochondria were prepared from SOL and AT
and were stored at –70°C until analysis (within 1 week). Mitochondria were
extracted in 50 mmol/l KH2PO4, 50 mmol/l K2HPO4, 10 mmol/l EGTA, 1 mmol/l
benzamidine, 50 µmol/l aprotinin, 50 µmol/l pepstatin, and 10 µmol/l leupeptin. Samples of mitochondrial extracts were denatured by heating to 60°C
with a 1:2 dilution of Laemmli electrophoresis buffer (0.25 mol/l Tris, pH 6.8,
10% glycerol, 0.01% bromophenol blue, 2% -mercaptoethanol, 2% SDS,
0.01 mol/l dithiothreitol). A total of 3 µg mitochondrial protein was then separated by discontinuous SDS-PAGE electrophoresis and was subsequently
transferred electrophoretically to nitrocellulose membranes with the electrophoresis semi-dry apparatus. Nitrocellulose filters were then blocked
overnight at 4°C with Tris-buffered saline (TBS) (150 mmol/l NaCl, 10 mmol/l
Tris-HCl, pH 7.6) supplemented with 0.05% Tween 20 and 5% (wt/vol) nonfat
powdered milk, incubated for 3 h at room temperature with polyclonal antisera raised against specific recombinant PDK isoforms, washed with TBS
with Tween (TBST; 0.05% Tween 20 in TBS) (3 5 min), and incubated with
horseradish peroxidase–linked secondary antibody IgG anti-rabbit (1:2,000 in
1% [wt/vol] nonfat milk in TBST) for 2 h at room temperature. The blots were
then extensively washed in TBST, and bound antibody was visualized with ECL.
The blots were then exposed to Hyperfilm (Amersham International, Buckinghamshire, U.K.), and the signals were quantified by scanning densitometry
and were analyzed with Molecular Analyst 1.5 software (Bio-Rad, Hemel
Hempstead, Hertfordshire, U.K.).
Statistical analysis. Experimental data are means ± SE. The statistical significance
of differences between groups was assessed with Student’s unpaired t test. Statistical comparisons were made with StatView (Abacus Concepts, Berkeley, CA).
Curve fitting was carried out with Fig P software (Biosoft, Cambridge, U.K.).
RESULTS
Food intake and body weight. Food intake and body weight
during the 4-week study period are shown in Fig. 1. Caloric
intake was ~65% higher in the high fat–fed group compared with
the control group. Similar increases in caloric intake in
response to high-fat feeding have been observed in previous
studies by other researchers (25). The increase in caloric intake
was not associated with an increase in body weight during the
4-week period of high-fat feeding. Although we did not systematically measure physical activity, no obvious differences
were evident in physical activity between the 2 dietary groups.
Effect of high-fat feeding on insulin action in vivo.
Insulin infusion increased plasma insulin concentrations from
12 ± 2 (n = 13) to 82 ± 5 µU/ml (n = 16) in the control rats and
from 13 ± 1 (n = 11) to 73 ± 4 µU/ml (n = 12) in the high fat–fed
rats. Neither the basal nor the clamped insulin concentrations differed significantly between the control and high
fat–fed groups. Basal glucose concentrations were 4.1 ± 0.1
(n = 13) and 3.8 ± 0.1 mmol/l (n = 11), respectively, in the control and high fat–fed groups. During hyperinsulinemia, glucose
was infused to maintain blood glucose concentrations, and
steady-state glucose concentrations were 4.2 ± 0.2 (n = 16) and
4.0 ± 0.1 mmol/l (n = 12), respectively, in the control and high
fat–fed groups. The Rd was significantly reduced (by 26%, P <
0.05) by high-fat feeding (Fig. 2A), which indicates the existence of peripheral insulin resistance. In control rats, glucose
utilization (transport + phosphorylation) rates (GUI values)
were significantly higher in SOL muscle than in AT muscle both
in the basal state and during hyperinsulinemia (by 5.9- and 3.4fold, respectively; P < 0.001). In addition, the response to
insulin infusion was 2.2-fold greater in SOL muscle than during hyperinsulinemia in AT muscle (Fig. 2B). This pattern is typical of slow-twitch versus fast-twitch muscle (11). GUI values
DIABETES, VOL. 49, MAY 2000
A
B
FIG. 1. Body weight and daily energy intake of rats maintained on control or high-fat diets. Rats were maintained on either a standard diet
() or a high-fat diet (). Data for body weight (A) and daily energy
intake (B) are means ± SE for 8–10 rats. No statistically significant
effects of high-fat feeding on body weights were evident. Effects of
high-fat feeding on daily energy intake were significant at all of the
time points studied (P < 0.05).
were significantly reduced by high-fat feeding in both the
basal state (by 54%, P < 0.01) and during hyperinsulinemia in
SOL muscle (by 44%, P < 0.01) and during hyperinsulinemia
in AT muscle (by 42%, P < 0.05). The mean increment in glucose utilization induced by hyperinsulinemia was reduced by
33% in SOL muscle and by 49% in AT muscle after high-fat feeding. We can conclude that the high-fat feeding protocol
induced insulin resistance regarding glucose utilization in
both muscle types in vivo.
Interconversion of active and inactive forms of PDC in
slow-twitch and fast-twitch skeletal muscle mitochondria.
Skeletal muscle PDC is converted from the inactive phosphorylated form to the active dephosphorylated form
(PDHa) by incubation of skeletal muscle mitochondria in
the absence of respiratory substrates and with respiratory
inhibitors (26). Conversely, incubation of mitochondria with
respiratory substrates increases the ATP concentration, thus
allowing phosphorylation of PDC by endogenous PDK.
Skeletal muscle mitochondria were prepared from control
rats in the fed state. Total PDC activities measured in the
absence of substrate and in the presence of respiratory
inhibitors and expressed relative to the mitochondrial
marker citrate synthase (see RESEARCH DESIGN AND METHODS)
were similar in mitochondria prepared from SOL and AT
muscles in control rats that were fed a standard diet ad libitum (Table 1). The administration of the high-fat diet produced
no significant change in total PDC activity in either type of
skeletal muscle. In contrast, high-fat feeding reduced the
percentage of active PDC in mitochondria from SOL muscle
incubated with respiratory substrate (2-oxoglutarate/
L-malate) from 37.6 ± 4.6% of total PDC in mitochondria from
control rats to 14.7 ± 1.5% of total PDC in mitochondria from
high fat–fed rats (P < 0.05). Because total PDC activity in SOL
mitochondria was unchanged by high-fat feeding, the effect
of high-fat feeding to lower the percentage of active PDC is
a consequence of increased net phosphorylation of PDC. By
contrast, the percentage of active PDC in mitochondria from
777
PDK ISOFORM EXPRESSION
A
B
FIG. 2. Whole-body Rd and GUI values in SOL and AT
muscles in the basal state and during euglycemiahyperinsulinemia in control and high fat–fed rats.
Whole-body Rd (A) and GUI (B) values were measured
in the basal (postabsorptive) state (, both control and
high fat–fed rats) or after 2.5 h of euglycemia-hyperinsulinemia in control ( ) or high fat–fed () rats.
Data are means ± SE for 5–12 rats. *P < 0.05 vs. control;
**P < 0.01 vs. control.
the AT muscle of high fat–fed rats incubated with respiratory
substrate remained similar to that found with mitochondria
from control rats (control rats 29.3 ± 4.2% of total PDC, high
fat–fed rats 24.2 ± 3.2% of total PDC).
Effect of high-fat feeding on PDK activities in slowtwitch and fast-twitch skeletal muscle. In control rats, no
significant differences in PDK activity were evident between
SOL and AT muscles (Table 1). PDK activity in SOL mitochondria was increased 2.2-fold (P < 0.01) by high-fat feeding
(Table 1). PDK activity in AT mitochondria was also significantly increased (P < 0.001) in response to high-fat feeding
(Table 1). The fold increase in PDK activity observed in AT
muscle (2.1-fold) was approximately similar to that observed
in SOL muscle (Table 1).
Protein expression of individual PDK isoforms in
skeletal muscle mitochondria. Western blot analysis was
used to determine whether the increased skeletal muscle
PDK activities elicited by high-fat feeding were because of
increased expression of PDK2 and PDK4 (the 2 PDK isoforms of which mRNA is expressed in significant amounts in
rat gastrocnemius muscle) (4). High-fat feeding caused a
significant 1.96-fold (P < 0.001) increase in the amount of
PDK4 protein expressed in SOL mitochondria (Figs. 3 and 4).
In contrast, high-fat feeding elicited only a modest (1.29-fold,
P < 0.01) increase in the amount of PDK2 protein present in
SOL muscle (Figs. 3 and 4). Thus, the relative expression of
PDK4 to PDK2 increases in SOL muscle in response to highfat feeding. High-fat feeding led to a 1.68-fold increase (P <
0.05) in PDK4 protein expression and to a 1.50-fold increase
(P < 0.05) in PDK2 protein expression in AT muscle (Figs. 3
and 4). Thus, the relative expression of PDK4 and PDK2 protein in AT muscle is relatively unchanged by high-fat feeding.
Effects of pyruvate on the percentage of PDHa in
skeletal muscle mitochondria from control and high
fat–fed rats. In mitochondria incubated with respiratory
substrate, pyruvate addition increases the percentage of
active PDC through suppression of PDK activity (assessed by
incorporation of [32P] from [32P]-labeled inorganic phosphate
in mitochondria incubated with 2-oxoglutarate/L-malate)
(26). The percentage of PDHa in respiring mitochondria from
SOL and AT muscles of control rats increased progressively
when the pyruvate concentration was increased successively
from 0.01 to 10 mmol/l (Fig. 5). Increasing the pyruvate concentration further to 100 mmol/l did not lead to further activation of PDC (data not shown). In control rats, the pyruvate
concentrations giving 50% PDHa were ~0.25 and ~0.51 mmol/l
for mitochondria prepared from SOL and AT muscles, respectively (Fig. 5). The pyruvate concentration giving 50% PDHa
in SOL muscle was increased 7.6-fold (to ~1.9 mmol/l) in
response to high-fat feeding. As a consequence, the percentage of active PDC in SOL mitochondria incubated with physiological pyruvate concentrations (0.01, 0.1, and 1 mmol/l) was
significantly reduced by high-fat feeding (by 61, 66, and 37%,
respectively; P < 0.05) (Fig. 5). In AT muscle, the effect of highfat feeding in decreasing the sensitivity of PDK to inhibition
by pyruvate was more modest compared with that in SOL
muscle, with only an ~3.9-fold increase in the pyruvate concentration giving 50% PDHa (to ~2.0 mmol/l) (Fig. 5). Thus,
TABLE 1
Total PDC and PDK activities in SOL and AT muscle from control and high fat–fed rats
Total PDC activity (mU/U citrate synthase)
Control
High fat
SOL
AT
182.0 ± 15.8 (5)
178.1 ± 7.4 (5)
175.3 ± 5.7 (4)
162.1 ± 29.0 (4)
PDK activity (min–1)
Control
0.348 ± 0.023 (5)
0.385 ± 0.029 (6)
High fat
0.761 ± 0.094 (8)*
0.789 ± 0.083 (14)†
Data are means ± SE for the numbers of preparations from individual rats shown in parentheses. Each assay was run in duplicate.
PDHa activities were measured in freeze-clamped muscle extracts. Total PDC activities were measured in extracts of freshly isolated skeletal muscle mitochondria incubated for 15 min in the absence of respiratory substrate. PDK activities were measured in
mitochondrial extracts. Rate constants for PDK activity were calculated by least-squares linear regression analysis of ln[% of 0 time
activity] against time. *P < 0.01 vs. control: †P < 0.001 vs. control.
778
DIABETES, VOL. 49, MAY 2000
M.J. HOLNESS AND ASSOCIATES
A
B
FIG. 3. Effects of high-fat feeding on PDK2 and PDK4 protein expression in mitochondria prepared from SOL and AT muscles of control and
high fat–fed rats. Rabbit polyclonal antisera raised against PDK2 and
PDK4 were used to detect these proteins with Western blot analysis.
Typical immunoblots of PDK2 (A) and PDK4 (B) protein expression
are shown for SOL and AT muscles of control and high fat–fed rats.
Muscle mitochondrial extracts were denatured and subjected to SDSPAGE and immunoblotting with these isoform-specific antibodies as
described in RESEARCH DESIGN AND METHODS. Each lane corresponds to
3 µg of mitochondrial protein. A total of 5–9 preparations of mitochondria were analyzed. Representative results are shown.
as shown graphically in Fig. 6, high-fat feeding was generally
less effective in impairing the pyruvate sensitivity of PDK in
mitochondria prepared from fast-twitch AT muscle than that
from slow-twitch SOL muscle.
FIG. 4. Quantification of Western analysis of PDK isoform expression
in extracts of mitochondria from the SOL and AT muscles of control
and high fat–fed rats. Western blots were analyzed by scanning densitometry using Molecular Analyst 1.5 software. Data are means ± SE
for 5–9 individual internally controlled experiments. *P < 0.05 vs.
control; ***P < 0.001 vs. control.
cles, with a more marked response in the slow-twitch muscles (27). Mitochondria prepared from SOL and AT muscles
were incubated with respiratory substrate and varying pyruvate concentrations to determine whether differences in the
sensitivity of PDK to inhibition by pyruvate were evident
between slow-twitch and fast-twitch skeletal muscles and
whether the selective changes in PDK isoform expression patterns evoked in these distinct skeletal muscle types in
response to high-fat feeding are associated with differential
DISCUSSION
A
In the present study, Western blot analysis with antibodies
raised against recombinant PDK2 and PDK4 conclusively
demonstrates that PDK2 and PDK4 proteins are expressed in
both slow-twitch (SOL) and fast-twitch (AT) skeletal muscle
in the fed state. The sensitivity of PDK to suppression by
pyruvate was found to be ~2-fold greater in SOL than in AT
muscle, which mirrors the increased insulin sensitivity that
is characteristic of slow-twitch versus fast-twitch muscle
(13). Furthermore, stable increases in PDK activities evoked
in response to high-fat feeding in both skeletal muscles are
associated with increased protein expression of 1 or both PDK
isoforms. The increase in PDK activity observed in response
to high-fat feeding in SOL muscle (the representative slowtwitch muscle) is mainly because of targeted upregulation of
PDK4. However, the enhanced PDK activity in the fast-twitch
muscle observed after high-fat feeding can be attributed to
increased protein expression of both PDK2 and PDK4.
To assess the physiological implications of muscle fiber–
specific changes in PDK isoform protein expression, we
examined the regulation of PDK activity by pyruvate. In vivo,
pyruvate may be derived from the glycolytic pathway or from
circulating lactate and alanine. Studies with recombinant
PDK isoforms have shown that PDK2 and PDK4 are differentially sensitive to inhibition by dichloroacetate, a pyruvate
analog (4). In vivo, the effectiveness of dichloroacetate to activate PDC differs between slow-twitch and fast-twitch mus-
FIG. 5. Effect of high-fat feeding on inhibition of PDK activity by
pyruvate in freshly isolated mitochondria from SOL and AT muscles
of control and high fat–fed rats. Freshly isolated mitochondria
(0.5–1.0 mg mitochondrial protein) from SOL (A) and AT (B) muscles
of rats maintained on a control diet () or a high-fat diet () were
incubated at 30°C with respiratory substrate (2-oxoglutarate/
L-malate) and the concentrations of pyruvate indicated as described
in RESEARCH DESIGN AND METHODS. Mitochondria were precipitated by
centrifugation, and steady-state PDC activity was measured in mitochondrial extracts. Data for 18 (control) and 6 (high fat–fed) mitochondrial preparations from individual rats are means ± SE for each
pyruvate concentration (concn.). *P < 0.05 vs. control.
DIABETES, VOL. 49, MAY 2000
B
779
PDK ISOFORM EXPRESSION
A
B
FIG. 6. A comparison of the sensitivity of PDK activity to suppression
by pyruvate in freshly isolated mitochondria from the SOL and AT muscles of control and high fat–fed rats. The results of the effects of
pyruvate on PDK activity in freshly isolated mitochondria from the
SOL (A) and AT (B) muscles of rats maintained on a control diet ()
or a high-fat diet () are expressed as percentages of the maximal
response to pyruvate observed with each individual mitochondrial
preparation. Further details are given in the legend for Fig. 5. Data for
18 (control) and 6 (high fat–fed) mitochondrial preparations from individual rats are means ± SE for each pyruvate concentration (concn.).
*P < 0.05 vs. control.
alterations in the sensitivity of PDK to inhibition by pyruvate
in vivo. In control rats, inhibition of PDK activity by pyruvate
was greater in mitochondria from slow-twitch muscle than in
mitochondria from fast-twitch skeletal muscle. Furthermore,
a more pronounced effect of high-fat feeding to reduce the
sensitivity of PDK to inhibition by pyruvate was observed in
slow-twitch skeletal muscle, whereas the effect of high-fat
feeding in fast-twitch skeletal muscle was relatively modest.
We therefore conclude that the pyruvate sensitivity of PDK in
skeletal muscle varies both with muscle fiber type (slowtwitch vs. fast-twitch muscle) and with nutritional status
(control vs. high fat–fed rats). In addition, important differences in PDK responses between skeletal muscles differing
in fiber composition can be revealed by the administration of
a diet high in saturated fat, which induces insulin resistance
at the level of glucose uptake and phosphorylation in skeletal muscle. In SOL muscle (the representative slow-twitch
muscle), the findings of concomitant changes in PDK activity and PDK4 expression together with decreased pyruvate
sensitivity are compatible with a functional switch to the
higher–specific activity less–pyruvate-sensitive PDK isoform
(PDK4) after high-fat feeding.
A previous study in vitro has shown that the presence of an
FA and dibutyryl cAMP in culture is necessary to maintain differences in PDK activities between freshly prepared SOL
strips from fed and starved rats (28). The data are consistent
with the concept that relatively high PDK activities are maintained in slow-twitch oxidative skeletal muscle when FA oxidation rates are high. The response of PDK4 expression to
increased dietary lipids is higher in SOL than in AT muscle;
this response parallels the propensity with which these muscles use lipid-derived fuels as respiratory substrates. In addition, given the specificity of the PDK isoform response to high780
fat feeding in SOL muscle (selective upregulation of PDK4),
our data suggest that FAs (or their metabolites) or metabolic
changes secondary to increased FA oxidation (e.g.,
decreased glucose utilization) are necessary for or facilitate
increased expression of PDK4 in slow-twitch oxidative skeletal muscle. In the present study, the high fat–fed rats exhibited a marked increase in caloric intake, but no increase in
body weight was observed (Fig. 1). A recent study by Mollica
et al. (29) demonstrated that a significant increase in energy
intake but no increase in body weight in high fat–fed rats
resulted from a significant increase in energy expenditure.
Skeletal muscle homogenates from these animals showed a
marked increase in FA-stimulated oxygen consumption,
which led to the suggestion that increased skeletal muscle FA
oxidation rates prevented excess fat deposition (29). Support
for a direct role of FA in signaling increased PDK4 expression
in skeletal muscle has recently been obtained. Feeding rats
WY14643, which is a very potent and selective agonist for
PPAR- (30,31), mimics the effects of starvation and diabetes to increase PDK4 expression in gastrocnemius muscle
(11). Increasing evidence exists that long-chain FAs or their
metabolites function as naturally occurring activators for
PPAR- (31–35). Thus, conceivably, the increased rates of FA
oxidation observed in skeletal muscle together with possible
direct effects of FAs on PPAR- underlie the increase in
PDK4 expression in skeletal muscles of high fat–fed rats.
However, we cannot infer whether FAs (or their metabolites) directly upregulate PDK4 expression in skeletal muscle
or whether high-fat feeding prevents downregulation of
PDK4 expression by inducing insulin resistance in skeletal
muscle. However, the degree of insulin resistance at the level
of glucose uptake and phosphorylation evoked by high-fat
feeding was similar in both SOL (slow-twitch) and AT (fasttwitch) skeletal muscle (15). Thus, our data imply that a positive correlation exists between increases in PDK4 expression
and the propensity with which different skeletal muscle
types use lipid-derived fuels as respiratory substrates but
not with the degree of insulin resistance induced by high-fat
feeding in individual skeletal muscles.
Exposure of fed rats maintained on a standard (high-carbohydrate/low-fat) diet to a high concentration of dichloroacetate leads to almost complete PDC activation in a range of
skeletal muscles within 2 h (27). This finding indicates that
PDK is active even in the fed state. Furthermore, it suggests
that regulation of skeletal muscle PDK by changes in the pyruvate supply is likely to be important in the regulation of
PDC phosphorylation status. Physiological conditions that
increase glycolytic flux in slow-twitch skeletal muscle (e.g.,
refeeding or exercise) would therefore be expected to
increase skeletal muscle pyruvate levels and suppress PDK,
thereby facilitating PDC activation and pyruvate oxidation. Our
data indicate that the specific increase in the protein expression of PDK4 induced by high-fat feeding in SOL muscle is
associated with a stable modification in the characteristics of
regulation of PDK by pyruvate such that activation of PDC secondary to inhibition of PDK by pyruvate is greatly impaired.
This finding suggests that differences in the regulatory characteristics of individual PDK isoforms demonstrated with
recombinant proteins in vitro (4) are relevant in vivo. The
loss of sensitivity of skeletal muscle PDK to suppression by
pyruvate observed after high-fat feeding may facilitate the
direction of glycolytically derived pyruvate toward lactate
DIABETES, VOL. 49, MAY 2000
M.J. HOLNESS AND ASSOCIATES
output rather than oxidation. During starvation, such an adaptation would be considered to be beneficial because pyruvate (and related 3C intermediates) could be released into the
blood and used for glucose synthesis by the liver for use by the
brain (36). However, this adaptation is potentially detrimental to glucose homeostasis after high-fat feeding because redirection of glycolytically derived pyruvate from oxidation
toward lactate output by skeletal muscle would facilitate
hepatic glucose production from lactate and may contribute
to the overproduction of glucose and ultimately the development of hyperglycemia. Evidence that an increase in the supply of gluconeogenic precursors from skeletal muscle may contribute to overproduction of glucose by the liver has been
obtained in an animal model of type 2 diabetes (low-dose
streptozotocin treatment) (37). Lactate production by perfused hindlimbs is significantly (3- to 4-fold) greater in diabetic
rats than in control rats (either perfused at normal glucose levels or at diabetic glucose levels) (37). Targeted pharmacological inhibition of the expression and/or activity of PDK4 in
slow-twitch muscle may therefore prove to be of considerable
importance as a strategy to prevent or ameliorate hyperglycemia associated with insulin resistance.
ACKNOWLEDGMENTS
This study was supported in part by grants from the British
Diabetic Association (RD98/1625) and the British Heart
Foundation (PG98/044) to M.C.S. and M.J.H., a grant from the
European Commission (Biomed 2 Programme) (BMH4CT97-2717) to M.C.S., and a grant from the National Institutes
of Health (DK47844) to R.A.H.
We thank Mr. Harjinder S. Lall for technical assistance.
REFERENCES
1. Sugden MC, Holness MJ: Interactive regulation of the pyruvate dehydrogenase
complex and the carnitine palmitoyltransferase system. FASEB J 8:54–61, 1994
2. Sugden MC, Holness MJ: Hormonal and nutritional modulation of PDHC
activity status. In Alpha-Keto Acid Dehydrogenase Complexes. Patel MS,
Roche TE, Harris RA, Eds. Basel, Birkhauser Verlag, 1995, p. 163–176
3. Randle PJ, Priestman DA, Mistry S, Halsall A: Mechanisms modifying glucose
oxidation in diabetes mellitus. Diabetologia 37 (Suppl. 2):S155–S161, 1994
4. Bowker-Kinley MM, Davis WI, Wu P, Harris RA, Popov KM: Evidence for existence of tissue-specific regulation of the mammalian pyruvate dehydrogenase
complex. Biochem J 329:191–196, 1998
5. Popov KM, Hawes JW, Harris RA: Mitochondrial alpha-ketoacid dehydrogenase kinases: a new family of protein kinases. Adv Second Messenger Phosphoprotein Res 31:105–111, 1997
6. Wu P, Sato J, Zhao Y, Jaskiewicz J, Popov KM, Harris RA: Starvation and diabetes increase the amount of pyruvate dehydrogenase kinase isoenzyme 4 in
rat heart. Biochem J 329:197–201, 1998
7. Sugden MC, Fryer LG, Orfali KA, Priestman DA, Donald E, Holness MJ: Studies of the long-term regulation of hepatic pyruvate dehydrogenase kinase.
Biochem J 329:89–94, 1998
8. Sugden MC, Holness MJ, Donald E, Lall H: Substrate interactions in the acute
and long-term regulation of renal glucose oxidation. Metabolism 48:707–715,
1999
9. DeFronzo RA, Jacot E, Jequier E, Maeder E, Wahren J, Felber JP: The effect of
insulin on the disposal of intravenous glucose: results from indirect calorimetry and hepatic and femoral venous catheterization. Diabetes 30:1000–1007, 1981
10. Majer M, Popov KM, Harris RA, Bogardus C, Prochazka M: Insulin downregulates pyruvate dehydrogenase kinase (PDK) mRNA: potential mechanism
contributing to increased lipid oxidation in insulin-resistant subjects. Mol
Genet Metab 65:181–186, 1998
11. Wu P, Inskeep K, Bowker-Kinley MM, Popov KM, Harris RA: Mechanism
responsible for inactivation of skeletal muscle pyruvate dehydrogenase complex in starvation and diabetes. Diabetes 48:1593–1599, 1999
12. Holness MJ, Sugden MC: Glucose utilization in heart, diaphragm and skeletal muscle during the fed-to-starved transition. Biochem J 270:245–249, 1990
13. James DE, Jenkins AB, Kraegen EW: Heterogeneity of insulin action in indiDIABETES, VOL. 49, MAY 2000
vidual muscles in vivo: euglycemic clamp studies in rats. Am J Physiol 248:
E567–E574, 1985
14. Ariano MA, Armstrong RB, Edgerton VR: Hindlimb muscle fiber populations
of five mammals. J Histochem Cytochem 21:51–55, 1973
15. Storlien LH, James DE, Burleigh KM, Chisholm DJ, Kraegen EW: Fat feeding
causes widespread in vivo insulin resistance, decreased energy expenditure,
and obesity in rats. Am J Physiol 251:E576–E583, 1986
16. Sugden MC, Holness MJ, Fryer LG: Differential regulation of glycogen synthase
by insulin and glucose in vivo in skeletal muscles of the rat. Am J Physiol
273:E479–E487, 1997
17. Fryer LG, Orfali KA, Holness MJ, Saggerson ED, Sugden MC: The long-term
regulation of skeletal muscle pyruvate dehydrogenase kinase by dietary lipid
is dependent on fatty acid composition. Eur J Biochem 229:741–748, 1995
18. Holness MJ, Sugden MC: Suboptimal protein nutrition in early life later influences insulin action in pregnant rats. Diabetologia 39:12–21, 1996
19. Sugden MC, Holness MJ: Modulation of in vivo insulin action by dietary protein during pregnancy. Am J Physiol 268:E722–E729, 1995
20. Ferre P, Leturque A, Burnol AF, Penicaud L, Girard J: A method to quantify glucose utilization in vivo in skeletal muscle and white adipose tissue of the anaesthetized rat. Biochem J 228:103–110, 1985
21. Caterson ID, Fuller SJ, Randle PJ: Effect of the fatty acid oxidation inhibitor
2-tetradecylglycidic acid on pyruvate dehydrogenase complex activity in
starved and alloxan-diabetic rats. Biochem J 208:53–60, 1982
22. Kerbey AL, Radcliffe PM, Randle PJ: Diabetes and the control of pyruvate dehydrogenase in rat heart mitochondria by concentration ratios of adenosine
triphosphate/adenosine diphosphate, of reduced/oxidized nicotinamide-adenine dinucleotide and of acetyl-coenzyme A/coenzyme A. Biochem J 164:509–
519, 1977
23. Kerbey AL, Randle PJ: Pyruvate dehydrogenase kinase/activator in rat heart
mitochondria, assay, effect of starvation, and effect of protein-synthesis
inhibitors of starvation. Biochem J 206:103–111, 1982
24. Priestman DA, Orfali KA, Sugden MC: Pyruvate inhibition of pyruvate dehydrogenase kinase: effects of progressive starvation and hyperthyroidism in
vivo, and of dibutyryl cyclic AMP and fatty acids in cultured cardiac myocytes.
FEBS Lett 393:174–178, 1996
25. Kraegen EW, James DE, Jenkins AB, Chisholm DJ, Storlien LH: A potent
effect of ciglitazone on muscle insulin resistance induced by high fat feeding
of rats. Metabolism 38:1089–1093, 1989
26. Fuller SJ, Randle PJ: Reversible phosphorylation of pyruvate dehydrogenase
in rat skeletal-muscle mitochondria: effects of starvation and diabetes.
Biochem J 219:635–646, 1984
27. Holness MJ, Liu YL, Sugden MC: Time courses of the responses of pyruvate
dehydrogenase activities to short-term starvation in diaphragm and selected
skeletal muscles of the rat. Biochem J 264:771–776, 1989
28. Stace PB, Fatania HR, Jackson A, Kerbey AL, Randle PJ: Cyclic AMP and free
fatty acids in the longer-term regulation of pyruvate dehydrogenase kinase in
rat soleus muscle. Biochim Biophys Acta 1135:201–206, 1992
29. Mollica MP, Iossa S, Liverini G, Soboll S: Stimulation of oxygen consumption
following addition of lipid substrates in liver and skeletal muscle from rats fed
a high-fat diet. Metabolism 48:1230–1235, 1999
30. Yu K, Bayona W, Kallen CB, Harding HP, Ravera CP, McMahon G, Brown M,
Lazar MA: Differential activation of peroxisome proliferator-activated receptors by eicosanoids. J Biol Chem 270:23975–23983, 1995
31. Forman BM, Chen J, Evans RM: Hypo lipidemic drugs, polyunsaturated fatty
acids, and eicosanoids are ligands for peroxisome proliferator-activated
receptors alpha and delta. Proc Natl Acad Sci U S A 94:4312–4317, 1997
32. Gottlicher M, Widmark E, Li Q, Gustafsson J-A: Fatty acids activate a chimera
of the clofibric acid-activated receptor and the glucocorticoid receptor. Proc
Natl Acad Sci U S A 89:4653–4657, 1992
33. Keller H, Dreyer C, Medin J, Mahfoudi A, Ozato K, Wahli W: Fatty acids and
retinoids control lipid metabolism through activation of peroxisome proliferator-activated receptor-retinoid X receptor heterodimers. Proc Natl Acad Sci
U S A 90:2160–2164, 1993
34. Gearing KL, Gottlicher M, Widmark E, Banner CD, Tollet P, Stromstedt M,
Rafter JJ, Berge RK, Gustafsson J-A: Fatty acid activation of the peroxisome
proliferator-activated receptor, a member of the nuclear receptor gene superfamily. J Nutr 124:1284S–1288S, 1994
35. Kliewer SA, Sunseth SS, Jones SA, Brown PJ, Wisely GB, Koble CS, Devchand
P, Wahli W, Wilson TM, Lenhard JM, Lehmann JM: Fatty acids and eicosanoids
regulate gene expression through direct interactions with peroxisome proliferator-activated receptors alpha and gamma. Proc Natl Acad Sci U S A 94:
4318–4323, 1997
36. Cahill GF Jr: Starvation in man. N Engl J Med 282:668–675, 1970
37. Mondon CE, Jones IR, Azhar S, Hollenbeck CB, Reaven GM: Lactate production and pyruvate dehydrogenase activity in fat and skeletal muscle from
control and diabetic rats. Diabetes 41:1547–1554, 1992
781